Botic Van De Zandschulp and Daniel Merida Aguilar meet in Shanghai on Wednesday, October 7, 2026, at 10:00 PM ET, and any Botic Van De Zandschulp vs Daniel Merida Aguilar prediction has to start with a contradiction: the matchup profile points one way, the market price points another, and the model ends up sitting on its hands anyway.
Merida Aguilar is priced at -120 on the moneyline, and the model's own read leans his way on paper. But the tier on this one is PASS, an official-card pass, which means no action is going out behind it. The reasoning is worth walking through anyway, because it is a genuinely close read, not a coin flip the model is dressing up.
Why the model leans Merida Aguilar
Over 201 service games against weaker servers on hard courts, Botic holds serve 85% of the time and breaks 25%. That sample exists because it is exactly the kind of opponent Merida profiles as: his serve carries just a 4.8% ace rate, well below a tour-standard weapon, and Botic's own first serve (9% aces) is built to keep return games away from his opponent's second ball. On raw serve-versus-serve terms, that split favors Botic.
The flip side is Botic's own second serve, which is the softest part of his game. The brief has him winning only 45-47% of his second-serve points and double-faulting on 5.5% of his service points. Merida's return has shown it can go after exactly that kind of number: he's won 53% of points against second serves generally, and 64% against Fucsovics's second serve specifically at the US Open. A returner winning 53% of second-serve points is clearing Botic's 45-47% second-serve win rate, and that gap is the single biggest reason the model gives Merida a look at all.
The case against
None of this is a clean case. Merida's most recent match was a clay-court Davis Cup win over Alejandro Tabilo, 18 days ago, useful mainly as a form signal rather than hard-court evidence given the surface change. His last hard-court match was 33 days back, a straightforward loss to Alex Michelsen at the US Open. Five weeks away from a hard court is five weeks of lost timing, and there is no newer hard-court result to say that rust is gone.
Botic, by contrast, has been on hard courts constantly. He won two qualifying matches at Beijing before losing a competitive three-setter to Francisco Cerundolo, a match in which he actually led after the first set and had break-point chances in the second. His only truly clean loss in the recent stretch was to Nikoloz Basilashvili, a player ranked well below him. Botic has also faced tougher competition on the whole than Merida has (the brief puts it at the 64th percentile versus the 51st), which cuts some of the shine off his recent record but doesn't erase the rust question hanging over Merida.
The number that matters
The 12-month hard-court numbers split in a way that doesn't fully settle it either. Botic holds serve 82% of the time and breaks 18%. Merida holds 79% and breaks 25%, though on a smaller sample. Merida's break rate sits above Botic's, and his hold rate sits a little below it. Add in that Merida's own serve rarely double-faults (3.7%, well under Botic's 5.5%), and the return-and-serve picture together nudges toward Merida without being decisive in either direction.
Put plainly: Botic's game is built to beat exactly the kind of weaker server Merida is, but Merida's return on second serve is built to beat exactly the kind of leaky second serve Botic has. Both things are true. That's why the model's own read calls this a thin lean toward Merida rather than a confident read on either player.
The verdict
This stays a PASS. The model leans toward Merida Aguilar on the matchup math, but it is not backing that lean with a bet, and no official action is going out on this one. That's not a comment on how close the read is, just a statement of what is and isn't being played.
- Event · Shanghai · ATP (2026) · Wednesday, October 7, 2026 · 10:00 PM ET
- Market · Daniel Merida Aguilar -120 — recorded when the pick was posted
- Model · T Money, Botcappers' AI Tennis model — how it is built and what it does not know
- Record · every pick is timestamped before the event and graded in public, win or lose
- Statistics · the model's own match/game database, current to the posting time above. Figures a reader cannot open here are internal to that database, not published third-party data.

